A new class “Modeling with statistics” started for undergraduate students at Yokohama National University.
In the class, I talk about practical data analysis skills for business. It is not just an academic statistics but how you find/create insights from data and apply them to practical solutions.
The main message of the class is “Goal-oriented” approach, where the thinking process is quite important. Many business persons/students tend to start to process data once received, but such approach would not show anything useful for the ultimate goal. In any case, you need to start with goal setting/definition and identify what data to collect and how to show/analyze the data in order to achieve the goal.
Main topic of this week was data aggregation/breakdown to effectively achieve the goal.
You need to think of the appropriate level of aggregation of data which many people tend to use the raw data without thinking about the level of aggregation.
One of the exercise in the class this week was as follows:
Students need to think of the most effective way to illustrate their messages by aggregating or breaking down the data. Also they are required to find the most effective graph to show the conclusion.
I gave them advice that “Outputs” are NOT necessarily your final “Message/conclusion”. You need to put the final conclusion, not just showing the graphs or tables.
As an output example, if you want to show the yearly trend by Product, then you may aggregate the regional data into the total by year and make a graph as follows:
As another example, if you want to show the sales ratio by region or by-product so that you can discuss the future potential market(s) or identify the area to focus, then you may illustrate the message as follows:
In this case, you can aggregate yearly data into the average to simplify the message.
Final example is as follows:
If you want to identify the potential market(s), then you may distinguish the trends across the markets. In this case, I would put more business focus on the Asian market rather than America.
This was a simple exercise, but the students struggled to define their goals(what to show) , to find an effective way to show the outputs and to summarize the conclusion briefly.
I believe this is a good exercise for them to learn how to utilize the data to create value in business operations. It must be a good skill for students to learn before going to the “real” business world!